US 20130147666A1

(19) United States (12) Patent Application Publication (10) Pub. No.: US 2013/0147666 A1
Abe et al.
(54) RADAR SYSTEM (52)

(43) Pub. Date:

US. Cl.

Jun. 13, 2013

CPC .................................... .. G01S 13/06 (2013.01)

(71) Applicant: DENSO CORPORATION, Kariya-city

(JP)
(72) Inventors: Yoshihiro Abe, Kariya-shi (JP);

USPC

........................................................ ..

342/451

Akiyoshi Mizutani, Nagoya (JP)

(73) Assignee: DENSO CORPORATION, Kariya-city

(57)

ABSTRACT

(JP)
(21) Appl. No.: 13/711,943
(22) Filed: Dec. 12, 2012

A radar system for detecting positions of targets using a high resolution algorithm is presented. The number of incorrect
target detections due to incorrect estimation of the number of radar re?ections is reduced. A two-dimensional (aZimuth and

distance) peak having maximum poWer is detected, and a

virtual beam formed centering on that peak. If that tWo dimensional peak is buried in the virtual beam it is assumed to be a noise peak and removed. This is repeated for all tWo

(30)

Foreign Application Priority Data

(JP) ............................... .. 2011-272475

Dec. 13, 2011

dimensional peaks, thereby removing noise peaks and reduc

ing the number of unnecessary target detections. The system
can also correctly detect peaks that are close together at a certain resolution but have a poWer larger than the virtual beam.

Publication Classi?cation

(51)

Int. Cl. G01S 13/06

(2006.01)

PERFORM FFT ANALYSIS

V EXTRACT DISTANCE-DIRECTION PEAKS
V

CALCULATE AZIMUTH

(MUSIC)
V

EXTRACT TWO-DIMENSIONAL PEAKS BASED ON DISTANCE AND AZIMUTH

V
PERFORM PAIR MATCHING

/ S1 50

V
GENERATE TARGET INFORMATION
v

/ S160

END

Patent Application Publication

Jun. 13, 2013 Sheet 1 0f 4

US 2013/0147666 A1

24

25

26

20
. FILTER

k A ~ g

k r

SIGNAL PROCESSOR

32

iTARGET
v /
INTER-VEHICLE

INFORMAHUN

/
ENGINE ECU % F

30

34

=* P Jii? U

/
BRAKE ECU =< F

Patent Application Publication

Jun. 13, 2013 Sheet 2 of4

US 2013/0147666 A1

V
PERFORM FFT ANALYSIS

/ S1 1 0

V
EXTRACT DISTANCE-DIRECTION PEAKS

/ S1 20

V
CALCULATE AZIMUTH

/ S1 30

(MUSIC) V

/ S1 40

EXTRACT TWO-DIMENSIONAL PEAKS BASED ON DISTANCE AND AZIMUTH

V
PERFORM PAIR MATCHING

/ S1 50

V
GENERATE TARGET INFORMATION
V

/ S1 60

END

Patent Application Publication

Jun. 13, 2013 Sheet 3 0f 4

US 2013/0147666 A1

E5 ,3
(EXTRACTION TWO DIMENSIONAL PEAKS) BASED ONOF DISTANCE AND AZIMUTH
V

EXTRACT TEMPORARY PEAKS

/ S220

SORT TEMPORARY PEAKS IN DESOENDING ORDER OF THEIR POWER

REGISTER TEMPORARY PEAK HAVING

MAXIMUM POWER As

S230

TWO~DIMENSIONAL PEAK

NO

/
FORM VIRTUAL BEAMS OF REGISTERED PEAKS

UNREMOVED '2

END

Patent Application Publication

Jun. 13, 2013 Sheet 4 0f 4

US 2013/0147666 A1

WIDTH CORRESPONDING TO AZIMUTH RESOLUTION

GAIN

m MUSIC SPECTRUM
---- VIRTUAL BEAM

NOISE PEAK

TWODIMENSIONAL PEAK TEMPORARY PEAKS

DISTANCE DIRECTION

VIRTUAL BEAMS

CONVENTIONAL RADAR

(USING MUSIC HIGH RESOLUTION ALGORITHM)

FREQUENCY 6.6%

EXEMPLARY EMBODIMENT

(MUSIC LVIRTUAL BEAMS)

1.8%

US 2013/0147666 A1

Jun. 13,2013

RADAR SYSTEM CROSS-REFERENCE TO RELATED APPLICATION

[0011] Hence present application presents a folloWing radar system as exemplary embodiment, of Which includes: i)
an electromagnetic Wave transmitting means for transmitting

electromagnetic Waves; ii) an electromagnetic Wave receiving

means for receiving electromagnetic Waves re?ected from

[0001]

This application is based on and claims the bene?t

of priority from earlier Japanese Patent Application No. 2011-272475 ?led Dec. 13, 2011, the description ofWhich is

incorporated herein by reference.

BACKGROUND OF THE INVENTION

targets; iii) a position estimating means for estimating a posi tion of each target that has re?ected the electromagnetic Waves, using at least a high resolution algorithm, according to

reception signals acquired from the electromagnetic Wave

receiving means; iv) a virtual beam creating means for creat ing virtual beams having a predetermined beam Width cen

[0002] [0003] detects rithm. [0004] [0005]

1. Technical Field of the Invention The present invention relates to a radar system that

tering on respective targets positions estimated by the posi

tion estimating means; and v) a removing means for removing
the targets that are located Within a range of the respective virtual beams and a received poWer With respect to the respec

the position of a target using a high-resolution algo

2. RelatedArt Some types of radar systems are Well knoWn. One

type of Well-knoWn radar system uses array antennas com

tive position is smaller than the poWer of the respective virtual beams (The ?rst aspect of the radar system). [0012] Accordingly, by means of the virtual beam, the ?rst
aspect of the radar system can remove the small poWer peak

posed of a plurality of antenna elements to detect the distance

to a target that has re?ected radio Waves or the direction of

Which appears in the vicinity of the peak regarding an angle

spectrum, that is, can decrease unnecessary detection of non

arrival of radio Waves (i.e. a direction in Which the target is

present).
[0006] As mentioned in JP-A-2006-047282, so-called

high-resolution algorithms, such as MUSIC (Multiple Signal Classi?cation), ESPRIT (Estimation of Signal Parameters via
Rotational Invariance Techniques) and the like, are Well knoWn as methods of estimating the direction of arrival of radio Waves. In a method using such a high-resolution algo

existent targets. Hence the present radar system can prevent causing unnecessary control of the vehicle. [0013] It is preferable that a Width of the virtual beam is just set to a value being able to gain a desired resolution (The second aspect of the radar system). According to the second aspect, the radar system can prevent over performance and offers a preferred detecting result to the doWnstream compo
nent for vehicle control.

rithm, an angle spectrum is prepared according to a correla tion matrix that shoWs correlation betWeen signals received by antenna elements (also referred to as channels). Then, in the method, the angle spectrum is scanned to estimate the
direction of arrival of radio Waves.

[0014] It is preferable that the removing means processes the estimated targets in descending order of their poWer (The third aspect of the radar system). According to the third
aspect, the processing load of the removing means can be reduced.

[0007] In a high-resolution algorithm, an angle spectrum is prepared in a process of calculating the direction of arrival of radio Waves. Speci?cally, in preparing an angle spectrum in
the process, the number of re?ections is estimated so that the

[0015] Though the position estimated by the estimating

means may be a one-dimensional position represented by aZimuth, a tWo -dimensional position represented by both dis tance and aZimuth is more preferable (The fourth aspect of the

estimated number of peaks is formed. Accordingly, if the

estimated number of incoming Waves is incorrect, unneces

radar system).
BRIEF DESCRIPTION OF THE DRAWINGS

sary peaks (noise peaks) may be formed to unnecessarily

detect a target that is not actually present (hereinafter this detection is referred to as unnecessary detection). An
unnecessary detection may cause unnecessary control such as

[0016] In the accompanying draWings:

[0017] FIG. 1 is a block diagram illustrating a con?guration
of a radar sensor applied to an ACC system, according to the

in an ACC (Adaptive Cruise Control) system of a vehicle. Therefore, the unnecessary detections have to be reduced as much as possible.

exemplary embodiment;
[0018] FIG. 2 is a ?oW diagram illustrating a processing performed by a signal processor in the radar sensor; [0019] FIG. 3 is a ?oW diagram speci?cally illustrating a tWo-dimensional peak extraction performed in the process
mg;

[0008]

When a high-resolution algorithm is used, peaks are

detected in the calculation With unnecessary ?ne resolution. Therefore, use of a high-resolution algorithm raises a prob

lem of exposing noise peaks that Would have been buried in

peaks based on the Waves re?ected from a target in so-called

Beam-former method. [0009] On the other hand, as described in JP-A-2006

047282 as Well, several attempts are made to enhance the

[0020]

FIG. 4 is an explanatory diagram illustrating an

outline of a virtual beam;

accuracy of estimating the number of re?ections. HoWever, currently, su?iciently satisfactory results are not yet avail
able.
SUMMARY

[0021] FIG. 5 is an explanatory diagram exemplifying posi tions of temporary peaks and gain distribution of virtual
beams on a tWo-dimensional map expressed by distance and

aZimuth; and
[0022] FIG. 6 is a table shoWing measurements of the
effects of the radar sensor.

[0010] Accordingly, regarding a radar system for detecting the position of a target using a high-resolution algorithm,
such a radar system that is able to reduce the occurring of
unnecessary detections of targets due to an erroneous estima

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0023]

With reference to the accompanying draWings,

tion of the number of incoming Waves is desired.

hereinafter is described an embodiment of the present inven

US 2013/0147666 A1

Jun. 13,2013

tion. FIG. 1 is a block diagram illustrating a con?guration of a radar sensor (corresponding to the radar system recited in

selects any one of the antenna elements composing the recep

tion antenna array 20 to supply reception signals (Sr) from the

selected antenna element to a doWnstream component. The

the claims) 1 applied to an ACC (Adaptive Cruise Control) system, according to the exemplary embodiment. [0024] As shoWn in FIG. 1, the ACC system includes the
radar sensor 1, an electronic control unit for inter-vehicle distance control (hereinafter referred to as an inter-vehicle

ampli?er 22 ampli?es the reception signals (Sr) supplied

from the reception sWitch 21. The mixer 23 mixes the recep

tion signals (Sr) ampli?ed by the ampli?er 22 With the local signals (L) to generate beat signals (BT). The ?lter 24
removes unnecessary signal components from the beat sig nals (BT) generated by the mixer 23. The A/D converter 25 samples the output from the ?lter 24 and converts the sampled output into digital data. The signal processor 26 activates or deactivates the oscillator 10, controls sampling of the beat

distance control EC ) 30: herein inter-vehicle distance

means distance betWeen the vehicles, an electronic control

unit for engine (hereinafter referred to as an engine ECU) 32 and an electronic control unit for brake (hereinafter
referred to as a brake ECU) 34. The radar sensor 1 detects

conditions ahead of the vehicle. The inter-vehicle distance control ECU 30 automatically controls inter-vehicle distance according to the conditions ahead of the vehicle, Which are detected by the radar sensor 1, and the conditions of the vehicle detected by various sensors, not shoWn. The engine ECU 32 controls the conditions of the engine according to a manipulated variable of the accelerator pedal or an instruction from the inter-vehicle distance control ECU 30. The brake ECU 34 controls the conditions of the braking device accord ing to the state of manipulation of the brake pedal or an instruction from the inter-vehicle distance control ECU 30. The ECUs 30, 32 and 34 are each mainly con?gured by a Well-known microcomputer and connected to each other via an on-vehicle LAN (Local Area Network). [0025] The radar sensor 1 is con?gured as a so-called mil
limeter-Wave radar based on an FMCW (Frequency-Modu

signals (BT) performed via the A/D converter 25, processes

the sampled data so as to generate target information or the

like.

[0029] Hereinafter, the reception systems using the antenna

elements CH1 to CHn composing the reception antenna array
20 are referred to as channels CH1 to CNn, respectively. The

signal processor 26, Which is mainly composed of a Well knoWn microcomputer, includes a processing unit (e.g., DSP (Digital Signal Processor)). The processing unit has a func tion, for example, of performing Fast Fourier Transform (EFT) processing With respect to data acquired via the A/D
converter 25.

[Operation of the Radar Sensor]

[0030] In the radar sensor 1 con?gured in this Way, When the oscillator 10 is activated according to the instruction from

lated Continuous Wave) system. The radar sensor 1 transmits and receives frequency-modulated electromagnetic Waves in
a millimeter-Wave band as radar Waves to and from capture

the signal processor 26, the high-frequency signals generated

in the oscillator 10 is transmitted to the ampli?er 12 for

targets, such as preceding vehicles or roadside objects (here

inafter referred to as captured targets). The radar sensor 1

ampli?cation. Then, the output poWer of the ampli?ed high frequency signals is distributed by the distributor 14 into the transmission signals (Ss) and the local signals (L). Of these
signals, the transmission signals (Ss) are transmitted as radar
Waves via the transmission antenna 16.

generates information regarding the captured targets (herein

after referred to as target information) and transmits the information to the inter-vehicle distance control ECU 30.

[0026] The target information at least includes positions (distance and aZimuth) of the captured targets and relative velocity of the captured targets With respect to the vehicle equipped With the radar system (hereinafter also referred to as a system-equipped vehicle).

[0031] The radar Waves emitted from the transmission antenna 16 and re?ected from a target are received by the

reception antenna elements composing the reception antenna array 20. In this case, only the reception signals (Sr) of a reception channel Chi (iIl to n) selected by the reception sWitch 21 are supplied to the ampli?er 22 for ampli?cation.

The reception signals (Sr) ampli?ed by the ampli?er 22 are

supplied to the mixer 23. Then, the mixer 23 mixes the recep

[Con?guration of the Radar Sensor]

[0027] The radar sensor 1 is speci?cally described beloW. The radar sensor 1 includes an oscillator 10, ampli?er 12, distributor 14, transmission antenna 16 and reception antenna

array 20. The oscillator 10 generates high-frequency signals

of a millimeter-Wave band. The high-frequency signals are
modulated so as to have a section in Which the frequency

tion signals (Sr) With the local signals (L) from the distributor 14 to generate the beat signals (BT), i.e. frequency compo nents equivalent to the difference betWeen the signals (Ss) and the signals (Sr). The beat signals (BT) are subjected to ?lter
ing in the ?lter 24 to remove unnecessary signal components. Then, the ?ltered signals are sampled by the A/D converter and the sampled signals as digital data are supplied to the signal processor 26. [0032] The reception sWitch 21 is ensured to sWitch the
channels CH1 to CHn such that all of the channels CH1 to

linearly increases With time (this section is hereinafter

referred to as a frequency-increase section) and a section in

Which the frequency linearly decreases With time (this section

is hereinafter referred to as a frequency-decrease section).

The ampli?er 12 ampli?es the high-frequency signals gener

ated by the oscillator 10. The distributor 14 distributes poWer

CHn are selected by a predetermined number of times during one modulation cycle of the radar Waves. Further, the A/D

of the output from the ampli?er 12 into transmission signals (Ss) and local signals (L). The transmission antenna 16 radi ates radar Waves according to the transmission signals (Ss).
The reception antenna array 20 is composed of an n number of

converter 25 is ensured to perform sampling in synchroniZa tion With the sWitching timing. Speci?cally, during one
modulation cycle of the radar Waves, sampled data are stored for each of the channels CH1 to CHn and for each of the

reception antenna elements (CH1 to CHn) that receive the

radar Waves.

frequency-increase and frequency-decrease sections of the

radar Waves.

[0028]

The radar sensor 1 also includes a reception sWitch

[Main Processing]
[0033] The main processing performed by the signal pro
cessor 26 is described along the How diagram illustrated in

21, ampli?er 22, mixer 23, ?lter 24, A/D converter 25 and signal processor 26. The reception sWitch 21 sequentially

US 2013/0147666 A1

Jun. 13,2013

FIG. 2. The main processing is repeatedly started on the basis

of one modulation cycle of the radar Waves as one measure

ment cycle.

step S230, among the temporary peaks, the one having a maximum poWer is registered as a tWo-dimensional peak and, at the same time, is removed from the list of the temporary

[0034] When the main processing is started, frequency

peaks.
[0042] At step S240, it is determined Whether or not any

analysis (EFT processing here) is performed ?rst, at step

S110, for the sampled data corresponding to one modulation cycle stored during one measurement cycle. During the fre quency analysis, the signal processor 26 calculates a poWer spectrum of the beat signals (BT) for each of the channels CH1 to CHn and for each of the frequency-increase and frequency-decrease sections of the radar Waves.

temporary peaks remain unremoved. If no temporary peaks remain unremoved (NO at step S240), the tWo-dimensional peak extraction is ended. On the other hand, if any temporary peak remains (YES at step S240), control proceeds to step S250. At step S250, virtual beams are created for the regis tered tWo-dimensional peaks. Then, at step S260, it is deter
mined Whether or not there are any temporary peaks buried in the virtual beams. [0043] FIG. 4 is an explanatory diagram illustrating an outline of a virtual beam. As indicated by the broken line in
FIG. 4, a virtual beam is created so as to have a beam shape

[0035]

Then, at step S120, the signal processor 26 searches

through the poWer spectrum calculated at step S110 to extract frequency components as peaks (hereinafter referred to as

distance-direction peaks). At step S130, the signal proces

sor 26 performs aZimuth calculation to obtain directions of arrival of the re?ected Waves that have caused the distance

direction peaks extracted at step S120, for each of the fre quencies and for each of the modulation sections of the dis

(beam Width) in Which gain is loWered by 6 dB at a position distanced from the tWo-dimensional peak to be processed, by
a half of a target aZimuth resolution (or distance resolution). An expression buried in a virtual beam refers to the received poWer of the temporary peak not being more than the

tance-direction peaks. Speci?cally, the signal processor 26 performs a high resolution algorithm, such as MUSIC, using
the same distance-direction peaks simultaneously detected in
the channels CH1 to CHn to obtain an angle spectrum.

poWer of the virtual beam at the position. The diagram of FIG.

4 is based on only a relationship With aZimuth. In the case of

[0036] At step S140, the signal processor 26 performs tWo dimensional peak extraction. In the tWo-dimensional peak
extraction, the signal processor 26 removes unnecessary

using a relationship With distance, the range of loWering by 6

dB is speci?ed not by aZimuth resolution but by range reso lution. OtherWise, a relationship With distance is treated simi
lar to a relationship With aZimuth.

noise peaks from the peaks (temporary peaks) Whose posi

tions are tWo-dimensionally expressed by the distance and
aZimuth obtained through steps S110 to S130, to extract peaks based on the Waves re?ected from the target (hereinaf
ter referred to as tWo-dimensional peaks).

[0044] FIG. 5 is an explanatory diagram exemplifying posi tions of temporary peaks and gain distribution of virtual
beams on a tWo-dimensional map expressed by distance and

[0037] At step S150, the signal processor 26 performs pair

matching. In the pair matching, the signal processor 26 com bines, among the tWo-dimensional peaks extracted at step S140 for every frequency-increase and frequency-decrease
section, the peaks based on the Waves re?ected from the same

aZimuth. Speci?cally, in FIG. 5, the positions of temporary peaks (shoWn by circled numerals l to 5 in a descending order
of the poWer of the peaks) are shoWn on the tWo-dimensional

target. Speci?cally, the signal processors 26 combines the tWo-dimensional peaks Which coincide With each other in the
signal levels of the distance-direction peaks as extracted at step S120 and in the directions of arrival as calculated at step
S130.

map expressed by distance and aZimuth. Also, for the tempo rary peaks having the three highest poWers, the gain distribu tion of virtual beams is shoWn in the map by grayscale. The gain distribution of each virtual beam appears, for example,
as shoWn in FIG. 5.

[0045] If no temporary peaks are buried in the virtual beams (NO at step S260), control returns to step S230 to repeat steps
S230 to S260. If any peaks are buried in the virtual beams

[0038] At step S160, the signal processor 26 calculates a distance and a relative velocity for each of the combinations
set at step S150, using a Well-knoWn process in an FMCW

(YES at step S260), control proceeds to step S270. At step

S270, the temporary peaks buried in the virtual beams are determined to be the peaks originated from noise and removed. [0046] At step S280, it is determined Whether or not there are temporary peaks that remain Without being removed. If

radar. Then, the signal processor 26 generates target informa tion according to the calculated distance and relative velocity
as Well as the aZimuth calculated at step S130 and then ends

the main processing. The target information generated in the

main processing is transmitted to the inter-vehicle distance
control ECU 30 for use in ACC or the like.

any temporary peaks remain unremoved (YES at step S280),

control returns to step S230 to repeat steps S230 to S280. If no

[Two-Dimensional Peak Extraction] [0039] The tWo-dimensional peak extraction is speci?cally

described along the ?oW diagram shoWn in FIG. 3. [0040] With the start of the tWo-dimensional peak extrac tion, the results obtained at step S130 (angle spectrum

temporary peaks remain unremoved (NO at step S280), i.e. if all of the temporary peaks have been processed, the tWo dimensional peak extraction is ended.

Advantages
[0047] As described above, in the radar sensor 1, the tem porary peak having the maximum poWer is registered as a tWo-dimensional peak. Then, a virtual beam is formed cen

obtained through MUSIC) is subjected to peak scanning, at

step S210, to thereby extract temporary peaks and prepare a list indicating tWo-dimensional positions and poWers of the temporary peaks. The number of temporary peaks equals to the number of incoming Waves, Which has been estimated during the process of MUSIC. [0041] Then, at step S220, the temporary peaks extracted at
step S210 are sorted in descending order of their poWer. At

tering on the position of the tWo-dimensional peak. Then, the temporary peak, if it is buried in the virtual beam, is removed
as a noise peak. The same process is repeatedly performed for

the temporary peaks remaining Without being removed, thereby ensuring extraction of tWo-dimensional peaks.
[0048] Thus, the radar sensor 1 is able to remove noise peaks in the event that the noise peaks are caused in the

US 2013/0147666 A1

Jun. 13,2013

vicinity of the tWo-dimensional peaks, and thereby avoid erroneously estimating the number of re?ections to be larger
than the actual number. Accordingly, the radar sensor 1 is able to prevent unnecessary detections. [0049] Further the radar sensor 1 is able to individually and separately detect a peak having a larger poWer than that of the virtual beam, in the event that the peaks are located close to each other With respect to a required resolution. [0050] FIG. 6 is a table shoWing measurements of the effects of the radar sensor 1. Speci?cally, the table of FIG. 6 shoWs measurements of the frequency of unnecessary detec

positions are one-dimensionally expressed (i.e., only by

either aZimuth or distance) as a result of using a high-resolu

tion algorithm.
What is claimed is:

1. A radar system comprising:

an electromagnetic Wave transmitting means for transmit

ting electromagnetic Waves;

an electromagnetic Wave receiving means for receiving electromagnetic Waves re?ected from targets; a position estimating means for estimating a position of

tions performed by a radar (using high resolution algorithm

(Music)) of conventional art and by the radar sensor 1 of the

each target that has re?ected the electromagnetic Waves, using at least a high-resolution algorithm, according to

received signals acquired from the electromagnetic

Wave receiving means; a virtual beam creating means for creating virtual beams

present embodiment (using combination of high-resolution

algorithm and virtual beams). The measurements are based
on a scenario that a certain vehicle running on the adjacent

having a predetermined beam Width centering on

lane passes the system-equipped vehicle. As shoWn in the table, the radar sensor 1 of the present embodiment has been con?rmed to reduce the number of unnecessary detections to 1/3 to 1/4 of those of the radar of conventional art.

respective targets positions estimated by the position

estimating means; and
a removing means for removing targets that are located Within a range of the respective virtual beams and a

[0051]

In the present embodiment, the components other

than the signal processor 26 of the radar sensor 1 correspond

received poWer With respect to the respective targets position is smaller than that of the respective virtual
beams. 2. The radar system according to claim 1, Wherein a Width
of the virtual beam is set to a value corresponding to a desired

to the electromagnetic transmitting/receiving means; the sig

nal processor 26 that performs steps S120 to S130 corre

sponds to the position estimating means; the signal processor 26 that performs steps S250 corresponds to the virtual beam creating means; and the signal processor 26 that performs
steps S260 to S270 corresponds to the removing means.

resolution. 3. The radar system according to claim 2, Wherein the removing means processes the estimated targets in descend

[Modi?cations]
[0052] An embodiment of the present invention has been
described so far. HoWever, the present invention is not limited to the above embodiment but may be implemented in various modi?cations in a range not departing from the spirit of the

ing order of their poWer. 4. The radar system according to claim 3, Wherein the position estimated by the position estimating means is a tWo dimensional position Which is de?ned by both aZimuth and
distance. 5. The radar system according to claim 1, Wherein the removing means processes targets Whose position is esti

present invention. [0053] The embodiment described above exempli?es that

the radar sensor 1 is applied to an ACC system. HoWever, this
should not impose a limitation. Further, as an alternative to

mated by the position estimating means in descending order

of their poWer.

MUSIC as a high-resolution algorithm used in aZimuth detec tion in the above embodiment, ESPRIT or other algorithms may be used.

[0054] In the embodiment described above, virtual beams are applied to the temporary peaks Whose positions are tWo

dimensionally expressed by distance and aZimuth. Alterna

tive to this, virtual beams may be applied to the peaks Whose

6. The radar system according to claim 1, Wherein the position estimated by the position estimating means is a tWo dimensional position Which is de?ned by both aZimuth and distance. 7. The radar system according to claim 2, Wherein the position estimated by the position estimating means is a tWo dimensional position Which is de?ned by both aZimuth and distance.